PROJECT SUMMARY Waterpipe (WP) tobacco smoking is prevalent among young adults in the United States with up to 40% of college students reporting waterpipe use in the past year. Its increasing popularity is partly caused by the misperception that waterpipe smoking is safe or at least less harmful than other forms of tobacco smoking. However, waterpipe mainstream smoke contains levels of harmful and potentially harmful constituents that are 10-100 times higher than those in cigarette smoke. The Food and Drug Administration (FDA) mandates that packages of WP tobacco carry a text-only message about the addictiveness of nicotine. However, the effectiveness of this mandate is limited because previous research suggests that graphic warning labels are more effective, and because WP smokers will not interact with the tobacco package in the social venues where most WP smoking is done. These limitations are addressed by the overarching objective of the parent grant to investigate the impact of the placement of text-only or graphic-with-text health warning labels on the WP itself on smoking behavior. The research proposed in this supplement will complement the parent grant by investigating novel approaches to modeling outcomes in Parent Grant (PG) Aims 1 and 2. In Supplement Aim 1, we will develop dynamic models for eye tracking data which will better inform optimal placement of warning labels on WPs. In Supplement Aim 2, we will develop multivariate models for smoking topography data which will identify effects of WP warning labels on smoking behavior that are ignored in standard single outcome models.